Polarforschung 68: 291 - 295, 1998 (erschienen 2000)

A New Computer Bathymetric Model of the Arctic Ocean

By Elena D. Daniel l and Sergey P. Maschenkov'

THEME 15: Geodynamics of the Arctic Region

Summary: A new digital bathymetric model of the Arctic Occan has becn cre­atcd in VNIIOkeangeologia under Civilian Research and Development Foun­dation grant as part of IOC/lASC/IHO project activity. This model for the firsttime incorporated abundant systematic analog Russian data which have beendigitized and merged with other available digital information. Processing pro­cedures required for compilation were developed specifically for individual ar­eas differing in bottom topography and input data.

INTRODUCTION

The paper presents a new bathymetric model covering the areaconsisting of the Deep Arctic Basin and surrounding continen­tal shelves, and the Norwegian-Greenland Basin of the NorthAtlantic. Such wide bathymetric coverage of the Arctic area isnecessary for presenting all interconnected structures of thisunique and varied part of the Earth lithosphere. Considerableamount of existing bathymetric data have been gathered duringsystematic surveys carried out by Russian researches fromVNIIOkeangeologia, Polar Marine Geosurvey Expedition(PMGE), Murmansk Arctic Geological Expedition and RussianNavy (GUNiO). The collection, editing, adjustment andgridding of these data and their merging with available westerndata using the modern computer technology has been accom­plished in VNIIOkeangeologia and resulted in a newbathymetric model of the Arctic Ocean. This model reliab1ycharacterizes the regional peculiarities of the ocean floor reliefand throws new light on its most important features in compari­son with earlier published srnall-scale maps and globalbathymetry compilations, it was proposed for use in interna­tional digital Arctic bathymetry project (BOCHAROVAet al. 1997,KOMARITZYN et al. 1997). New features ofthe ocean bottom ge­ometry and bathymetry identified from our model have beeninterpreted together with potential field data for the purpose ofstudying the lithosphere structure (DANIEL et al. 1997,ZAYONCHEK et al. 1998).

Input data

Digital data base of the Arctic Ocean bathymetry was developedusing three main types of input data (Fig. 1): (a) original ana-

I All Russia Research Institute for Geology and Mineral Resources of the World Ocean(VNIIOkeangeologia). 1 Angliysky pr., 190121 SI. Petersburg, Russ ia,<dani@vniio.nw.ru>, mascha@vniio.nw.ru

Manuscript received 19 January 1999, accepted 10 Ocrober 1999

log point observations acquired by the Russian expeditions anddigitally captured in the course of this compilation; (b) gridseither developed from GUNiO data under project supported byCivilian Research and Development Foundation (CRDF) 01'

obtained from public domain data; and (c) digitized contoursfrom published bathymetric maps.

Original analog observations conducted in the course of Rus­sian geological and geophysical investigations resulted in acqui­sition of three types of point data (Tab. 1). The methods of ob­servations, applied technical facilities, and accuracy of naviga­tion and bottom depth estimations were mainly chosen subjectto natural features of the study areas. The eastern part of theRussian Arctic Shelf and deep Arctic Basin were covered byaircraft supported surveys with density of landings ranging from10-15 to 30-40 km. The ice-free Barents and Kara Seas werestudied by systematic shipborne surveys with lines and stationsspaced at 10-20 km and 3-7 km, respective1y. Bottom depthsestimations and their adjustment were perforrned in accordancewith commonly accepted hydrographic standards. In the DeepArctic Basin depths measurements were acquired at 1-3 km in­tervals from "North Pole" drifting ice-camp stations (KISELEV1986). Additional information was provided by German RN"Polarstern" (FÜITERER 1992, 1994) in the form of raw digitaldata; despite continuous digital acquisition along extendedobservation lines, this set contained a large amount of randomerrors and was on1ypartly included into our compilation. Char­acterization of input digital data is given in Fig.1.

DATA PROCESSING

Data were processed using the software developed in GeologicalSurvey of Canada (VERHOEF et al. 1995) for digital compilationof potential fields and other geophysical data. The processingincluded the following succession of operations: 1) editing, cor­relation and preliminary gridding of Russian point observationsdata; 2) visual comparison of contours derived from Russiangridded point data with other available contours data and devel­oping of intermediate grids; 3) integration of all obtainedgridded data sets to a final grid.

Differences in field surveys methods resulted in variety of pro­cedures for editing of initial data. Observations of type 1 (Tab.1) in the areas of overlapping surveys were sorted out manuallyfor mutual comp1iance. In nearshore zones these observationswere additionally examined for conformity with World VectorShoreline (WVS) data provided by the software. Russian

291

SOUE

170"E

110"E

Original point observations

./"0"'"

/

I: + :1I I

I I~[[[[[]]]

from surveys by Russian air-borne and ship-borne expeditions

along "North Pole" drifting stations tracklines

along RV "Polarstern" tracklines (FUTTERER, 1992, 1994)

Gridded data

obtained under CRDF project at regular lOxlO km interval from Russian point data

from OAKEY et a1., 1994 (regular 5x5 km interval used only in area which no ohter data)

Digitized contours

from 1:2,313,000 bathymetric map (CHERKIS et a1., 1991)

from 1:3,000,000 bathymetric map (CHERKIS et al., 1994)

from 1:500,000 bathymetric map (MATISHOV et al., 1995)

Fig. 1: Location diagram for input data used in digital bathymetric model

shipborne data were adjusted by means of eross-over analysis.Coherent data were prepared by correcting the depth values tomake discrepancies not more than 3-5 m in the depth range 1essthan 100 m and not more than 10-15 m in depths greater than100 m. After correlation of all shipborne surveys data sets theroot mean-square error of the surveys conducted within thedepth range about 20-600 m accounted for 5 m. Observationsof type 3 (Tab. 1) and RJV "Polarstern" raw data were edited

292

by procedure of detection and removal of spikes using fourthorder difference calculation. After correlation all "North Pole"data were included in the data set, whereas RJV"Polarstern" datawere used main1y in deep sea areas. The edited and adjustedpoint data sets were gridded by Minimum curvature (SMITH &WESSEL 1990) and Kriging (CRESSIE 1990) methods using vari­ous grid intervals. The primary grid intervals secured by the realobservations density were from 3 to 15 km.

Type of observations Navigation (type, accuracy) Depth estimations

(method, accuracy)

1. Landings on ice (regular coverage) Radio Navigation - ±200-±600 m Single channel seismic 0.1 % - 1 % depth

in East-Siberian Sea, Laptev Sea, Integrated Radio Navigation Echosounding 0.5 - 1 % depth

Chukchi Sea, Arctic Basin and Astronomy ±600 m Lead ±5m

2. Systematic shipborne surveys Satellite ±300 m Echosounding 1 % depth

in Barents Sea, Kara Sea Radio Navigation ±200 m Lead±5m

Astronomy ±1000 m

3. Observations from "North Pole" Satellite ±300 m Single channel seismic 0.1 % - 1 % depth

drifting stations Astronomy ±1000 m

Tab. 1: Types of Russian bathymetric point observations in the Arctic Occan.

The next step was to test the contours derived from Russiangridded point data for compatibility with other availablebathymetric information. For that purpose the contours plottedfrom newly developed Russian grids were visually comparedwith published bathymetric maps (ATLAS OF THE OCEANS 1980,PERRY et al. 1986, CHERKIS et al. 1991, 1994, MATlSHOV et al.1995) and contours plotted from public domain grids (COAKEYet al. 1994, NOAA 1988). It appeared that, in comparison withisobaths obtained from new Russian grid, the data of ATLAS OFTHE OCEANS (1980), PERRY et al. (1986) and NOAA (1988) weremuch more generalized, whereas maps published by CHERKIS etal. (1991, 1994) and more detai1ed map of MATISHOV et al.(1995) demonstrated good agreement and therefore were usedin full extend for western Eurasian margin area. This also con­firmed reliability of our new grid and justified using it as thebasis for digital compilation in the entire area covered by Rus­sian point observations. Data of OAKEY et al. (1994) main1ybased on digital contours from GEBCO Sheet 5.17 (IONES 1994)which also existed in the same area were found less informa­tive, and for that reason this digital set was included in the fi­nal compilation only where no other evidence was available.

After selection of all input data sets it was necessary to developintermediate grids for separate seabed portions differing in prin­cipal features of bottom relief. The following specific seabedportions were recognized: (i) nearshore zone, (ii) western she1fand slope, (iii) eastern shelf and slope, (iv) Norwegian-Green­land Basin, (v) Gakkel Ridge, (vii) Alpha-Mendeleev Ridge,(viii) Chukchi Borderland, and, finally, (ix) deep water basinsof the Arctic Ocean.

In the nearshore zone straightforward compilation results inappearance of numerous artifacts caused by a jagged con­figuration of coastline. Hence it was found practicable to formthe specific 1.5 km grid in this zone from our input bathymetrydata supplemented by land component of NOAA (1988). It al­lowed to minimize the discrepancies between zero-depth gridvalues and WVS data. Deeper and more differentiated westernshelf/slope and Norwegian Greenland Basin were gridded us-

ing the minimum curvature method with various tension factorsat 2.5 km interval, in shallower and flatter eastern shelf/slopethe same method was applied with high tension factors. Becauseof differences in bottom topography and general geometry of thebathymetric highs in the Deep Arctic Basin (NARYSHKIN 1995)the respective data sets were processed using individual com­bination of grid parameters. In all cases directional gridding byKriging method was applied at the intervals between 3 and 10km which enabled to enhance the anisotropie appearance ofthese bottom features. Deep water basins were found sufficientlyuniform to allow similar processing procedure by minimumcurvature method at 10 km interval. The above specific gridswere merged in a single gridded data set which was subsequentlyintegrated with OAKEY et al. (1994) data to form a final 5 kmgrid.

DISCUSSION

A color contour image of the digital model of the Arctic Oceanrelief is presented in Figure 2. The color depth scale is chosento portray the main relief features, whereas additional contoursdisplay smaller details.

A new bathymetry compilation is the next stage of updating theinternational digital bathymetry of the Arctic Ocean. Our dig­ital model is based on much greater amount of source data thanearlier digital compilations, and in our view, offers the oppor­tunity to produce a computer print-out at 1 : 5 000 000 scalewhich for the first time will be truly backed by the quality anddensity of original information. For the Eurasian Arctic shelf thecreditability of input data is even higher and allows re1iablehard­paper visualization at 1 : 2500000 scale. The contro1 of trust­worthiness of our model is provided by its good agreement withnewly published manual 1 : 5 000 000 map (BOTTOM RELIEF OFTHE ARCTIC OCEAN 1999).

Many features of the Arctic Ocean seabed that are important forregional tectonic interpretations and geopolitical issues (such as

293

N\0+:-

50'E

10'W

\\

70'W

1l0'Edepth scale (m)

50 200 800 1500 2500 3800 4400additional contours at depths 40, 70, 1800,2200,3200,3400,4200 m

20 100 500 1000 2000 3000 4000 5000

Fig. 2: Digital color image of the bottom relief of the Arctic Ocean

r(

~'\i }

170'E

delimitation of the outer extent of the continental shelf in Lawof the Sea context) but only preliminary outlined in earlier com­pilations can now be recognized with greater confidence. Thisrefers in the first place to deep water portion of the Arctic ba­sin where coherent geometry of the Gakkel and LomonosovRidges and separating basins is combined with clearly expressedsegmentation of the Gakkel Ridge and apparent heterogeneityof the Lomonosov Ridge relief. Also noteworthy is the regularinterval (500-600 m) in bathymetric highs in the central part ofthe Arctic ocean. Some new essential details have also emergedin sub-polar part of the Canada Basin.

CONCLUSION

A new digital model of the Arctic Ocean bottom relief has beendeveloped in VNIIOkeangeologia and tested in comparison withboth earlier digital compilations and modern manual products.It has also been successfully applied in lithospheric studies basedon joint analysis of bathymetry and potential fields(MASCHENKOV et al. 1999). The model may be regarded as thefirst step in creating a comprehensive integration of digital Are­tic Ocean bathymetry currently undertaken by the Arctic earthscience community under IOC-IASC-IHO project (MACNAB &GRIKUROV 1997) for the purposes of improved cartographicimaging and resource management in the High Arctic.

ACKNOWLEDGMENTS

We extend our gratitude to OUf colleagues from MAGE,PMGPE, GUNiO, and VNIIOkeangeologia who for many yearshave been collecting the data used in this compilation. We arealso thankful to scientists from the Naval Research Laboratory(USA), Alfred Wegener Institute (Germany), Geological Sur­vey of Canada who made available digital data and software.We gratefully acknowledge the reviewers G. Grikurov and R.Macnab for their constructive and helpful comments on themanuscript. The work has been performed under grant NRG 1­224 founded by the Civilian Research and Development Foun­dation.

References

Atlas of the Oceans (1980): The Aretie Oeean floor relief map, sealeI :30,000,000.- In: S.G. GORSHKOV (ed.), The Aretie Oeean, volurne III,Head Department of Navigation and Oeeanography, 184 pp. (In Russian).

Bocharova, E., Daniel, E., Zavonchek, A.& Maschenkov, S. (1997): Computerderived bathymetry maps in the Northern Eurasian Shelf. Abst. NII.- In:R. MACNAB & G. GRIKUROV. Aretie Bathymetry Workshop. St.-

Petersburg Russia, September 18-19, 1997, GCS Open File 3569, Geologi­cal Survey of Canada, Dartmouth, NS.

Bottom ReliefofThe Aretie Ocean (1999): Head Department of navigation andOceanography- VNIIOkeangeologia. Seale 1:5,000,000. St.- Petersburg.

Cherkis, NZ., HS. Fleming, M.D., Max, PR., Vogt & Czarnecki, M.F. (1991):Bathymetry of the Barents and Kara Seas, scale I : 2,313,000.- GeologiealSoeiety of America, Boulder CO.

Cherkis, NZ. & Vogt, PR. (1994): Regional bathymetry of the Norwegian­Greenland Sea, scale 1:3,000,000.- Naval Research Laboratory, Washing­ton, D.C. 20375 USA.

Cressie, NA. C. (1990): The origins of Kriging.- Mathcmatical Geology, v. 22:239-252.

Daniel, E.D., Bocharova, E. V & Zayonchek, A. V. (1998): Morphostruetural fea­tures of the eentral Aretie Basin.- 6'hZonenshain Conferenee on Plate tec­tonics. Moskow, February 17-20, 1998. Programme & Abstracts: 64.

Fütteret;D.K (1992): Aretie'91: The expedition ARK- VIIV3 ofRV "Polarstern"in 1991.- Rep. Polar Res. 107: 267 pp.

Fiitterer, D.K. (1994): The Expedition ARCTIC '93: Leg IX/4 of RV"Polarstern" 1993.- Rep. Polar Res.149: 244 pp.

Iones, M.T. Tabor; AR. & Weatherall, P (1994): GEBCO Digital Atlas: CD­ROM and supporting volume.- British Oceanographie Data Centre,Birkenhead, UK. -

Kiselev, Yu.G. (1986): The deep geology of the Aretie Basin.- Nidra, Moseow:224 pp. (in Russian).

Komaritryn, A.A., Bocharova, E. V & Daniel, E.D. (1997): Computer derivedbathymetry map of the Central Aretie Basin, abst. N3.- In: R. MACNAB& G. GRIKUROV. Aretie Bathyrnetry Workshop. St.-Petersburg, Russia,September 18-19, 1997, GCS Open File 3569, Geologieal Survey ofCanada, Dartmouth, NS.

Macnab, R. & Grikurov, G. (1997): Arctic Bathyrnetry Workshop. St.­Petersburg, Russia, September 18-19,1997.- GCS Open File 3569, Geol.Surv. Canada, Dartmouth, NS.

Maschenkov, S.P, Brozena, .I.M., Kovacs, L.c., Glebovskiy, VYu., Zayonchek,A V, Childers, VA & Daniel, E.D. (1999): New compilation of potentialfield and bathymetry maps on the basis of joint digital proeessing of US andRF Data Sets in the High Seas Arctie //Eos, AGU-99.- Boston: S 193.

Matishov, G.G., Cherkis, NZ., vermillion, M.S. & Forman, SL (1995):Bathymetrie map ofthe Franz Josef Land area, scale I: 500,000.- Geologi­eal Soeiety of Ameriea, Bouder CO.

National Oceanographic and Atmospheric Administration. (1988): Digital re­lief of the Earth. Data Announeement 88-MGG-02.- NOAA, National Geo­physieal Data Center, Boulder CO.

Naryshkin, G.D. (1995): Orographie map of the Aretie Basin, sc ale I:5,000,000.- All Russia Research Institute for Geology and Mineral Re­sources ofthe World Oeean and Head Department ofNavigation and Ocea­nography Ministry of Defenee Russian Federation.

Oakey, G., vardy; D., Moir; P, USOlV, K, Verhoef, J. & Macnab, R. (1994):Bathymetry and topography, Aretie and North Atlantie ace ans and adjaeentland areas, seale I: 10,000,000.- Geologieal Survey of Canada Open File3282B, Dartmouth, NS.

Perry, R.K & Fleming, HS. (1986): Bathymetry of the Aretie Oeean. Map atseale 1:4,704,075.- Naval Research Laboratory, Washington, D.C.

Smith, W.HF. & Wessei, P (1990): Gridding with eontinuous eurvature splinesin tension.- Geophysies 55: 239-305.

Verhoef, J., Oakey, G., Roest, w., Usow, K & vardy, D. (1995): Potential fieldsdata proeessing. The software programs paekage.- GSC Atlantie, GeologiealSurvey of Canada, Dartmouth, NS, Blue Vajra Computing, Halifax, NS.

Zayonchek, A. V, Childers, VA. & Daniel. E.D. (1998): Joint analysis of on-icegravity and digital bathymetry data in the Amerasian basin.- III IntemationalConferenee on Aretie Margins (ICAM III). Abstracts: 201.

295